An clever and automatic anti-unsolicited mail framework is vital because of the excessive increase of unsolicited e-mail assaults and the inherent malevolent dynamic inside the ones assaults on numerous social, personal, and expert work. There is an increased risk of identity theft, theft of sensitive information, financial loss, damage to reputation, and other crimes that threaten the privacy of the victim. When taking into account the multidimensional feature set of email, current methods are rather fallible. We believe that an artificial intelligence-based strategy is the most effective one going forward, particularly unsupervised machine learning. Exploring the application of unsupervised learning for ham and spam clustering in the mail by comparing these Random Forest, Logistic, Random Tree, Bayes Net, and Naïve Bayes algorithms with LTSM Algorithms by using frequency weightage of words and validating the best accuracy is the purpose of this study.
Jaya et al. (Thu,) studied this question.